Beware that this post is related to an SDK in maintenance mode
More Info: Consider nRF Connect SDK for new designs

nRF528xx custom UARTE baudrates support

I'm referencing the post  UART BAUDRATE register values where the calculation for custom baudrates are given by an nordic support engineer.

Currently we are trying to merge custom baudrate support to Zephyr. Maintainers hesitate to accept the PR because there is no support for custom baudrates mentioned in the nRF528xx references.

Can one official nordic (support) engineer validate that using custom baudrates is totaly fine and within the specs of nRF51/nRF52/nRF53 UARTE?

Parents
  • Hi,

     

    That formula is very much an approximation, as already stated in the linked case.

    If you look at the register definitions and its values for NRF_UART vs. NRF_UARTE:

    https://infocenter.nordicsemi.com/topic/ps_nrf52840/uarte.html?cp=5_0_0_5_33_8_26#register.BAUDRATE

    https://infocenter.nordicsemi.com/topic/ps_nrf52840/uart.html?cp=5_0_0_5_32_9_22#register.BAUDRATE

    You can see that there are differences for the target baud rate, vs. the register value.

     

    Can one official nordic (support) engineer validate that using custom baudrates is totaly fine and within the specs of nRF51/nRF52/nRF53 UARTE?

    As stated in the linked case, this is an approximation for the NRF_UART peripheral (used on nRF51-series, and to some extent; possible to use on some nRF52-series devices as well), and you must verify if the output is matching your desired baudrate.

    The output baudrate will differ when running on other peripherals, such as the DMA capable NRF_UARTE peripheral.

     

    Are there specific baudrates that is required to be used by your application?

     

    Kind regards,

    Håkon

  • Hi Håkon,

    what exactly changed with NRF_UARTE compared to the nRF51 NRF_UART?

    We are using the baudrates 6250, 12500, 18750, 25000, 31250, 37500, 43750, 50000, 56250, 62500, 68750, 75000, 81250, 87500, 93750, 100000 and used the approximation to calculate the register values. All baudrates work perfectly fine. Our scope as well our µC on the other end understand the nRF52840 and nRF52833 perfectly fine at all tested baudrates with the exact patch from the pull request.

    Best regards,

    Benedikt

  • Hi,

     

    bamkrs said:
    what exactly changed with NRF_UARTE compared to the nRF51 NRF_UART?

    NRF_UARTE is a redesigned peripheral based on the NRF_UART with EasyDMA support.

    bamkrs said:
    We are using the baudrates 6250, 12500, 18750, 25000, 31250, 37500, 43750, 50000, 56250, 62500, 68750, 75000, 81250, 87500, 93750, 100000 and used the approximation to calculate the register values. All baudrates work perfectly fine. Our scope as well our µC on the other end understand the nRF52840 and nRF52833 perfectly fine at all tested baudrates with the exact patch from the pull request.

    This is good to hear. As mentioned, the formula has always been an approximation, and may not work for all calculated bauds "in between", depending on what nRF device is targeted.

    That being said, I am glad to hear that this works as expected in your scenario.

     

    Kind regards,

    Håkon

Reply
  • Hi,

     

    bamkrs said:
    what exactly changed with NRF_UARTE compared to the nRF51 NRF_UART?

    NRF_UARTE is a redesigned peripheral based on the NRF_UART with EasyDMA support.

    bamkrs said:
    We are using the baudrates 6250, 12500, 18750, 25000, 31250, 37500, 43750, 50000, 56250, 62500, 68750, 75000, 81250, 87500, 93750, 100000 and used the approximation to calculate the register values. All baudrates work perfectly fine. Our scope as well our µC on the other end understand the nRF52840 and nRF52833 perfectly fine at all tested baudrates with the exact patch from the pull request.

    This is good to hear. As mentioned, the formula has always been an approximation, and may not work for all calculated bauds "in between", depending on what nRF device is targeted.

    That being said, I am glad to hear that this works as expected in your scenario.

     

    Kind regards,

    Håkon

Children
No Data
Related